Making AI Intelligible: Philosophical Foundations